I paid attention to the details this time before deciding, and stopped assuming the obvious answer was right. The short answer is that how consistent match quality is week to week becomes obvious only after a month. That held reliably once the first fortnight was out of the way.

I ran this for a couple of months with a couple of accounts, so this is not theoretical. Realistically whether read status is shown consistently changes noticeably month to month. It was noticeable in a way I could measure long after the novelty had worn off. Take it as one data point in the end, depending on where you are.

I looked at this from a few angles across two cities, and it was worth the effort. More than anything, how much depth the profiles allow can be tested inside a week.

The handful of things that made a real difference:

  • Whether the privacy policy is readable — it decided things for me.
  • Find out how long before the first upsell prompt appears — it is easy to test and rarely advertised.
  • How many photos the format actually allows
  • Find out whether there is a desktop experience at all

I did this the slow way across two cities, which changed my answer completely. Across all of them, the honesty of the pricing page drives the outcome more than luck does. It was consistent to a degree that mattered after I controlled for the obvious variables.